INDEX

Explanations

phrases indicating conditions that lead to physical or regulatory consequences

oai_token-act-pair · gpt-4o-mini Triggered by @bot

foreign languages or specific technical terms

np_acts-logits-general · gemini-2.5-flash-lite

the neuron detects causal or purposive connectives—words and short phrases that signal reason or intent (e.g., "because", "in order to", "to", "in").

oai_token-act-pair · gpt-5-mini Triggered by @vetterc0

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 9-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

]();

-0.54

httphttps

-0.53

पया

-0.50

 Audiodateien

-0.48

 صوتيه

-0.48

ագրություններ

-0.44

 للاسماء

-0.43

 postIndex

-0.43

ValueStyle

-0.42

POSITIVE LOGITS

 valsty

0.39

 gemeint

0.38

 bulundu

0.38

 Aufnahme

0.38

wiegend

0.37

badlogic

0.36

 čierna

0.36

 condiv

0.35

tierrez

0.35

 Vordergrund

0.35

Activations Density 0.685%

phrases indicating conditions that lead to physical or regulatory consequences

foreign languages or specific technical terms

the neuron detects causal or purposive connectives—words and short phrases that signal reason or intent (e.g., "because", "in order to", "to", "in").

No Comments

No Known Activations

phrases indicating conditions that lead to physical or regulatory consequences

foreign languages or specific technical terms

the neuron detects causal or purposive connectives—words and short phrases that signal reason or intent (e.g., "because", "in order to", "to", "in").

No Comments

No Known Activations