INDEX

Explanations

evidential phrases indicating contrasts or comparisons

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

contrasts or comparisons

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

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

WireFormatLite

-0.47

 ſta

-0.46

السكان

-0.46

henswürdigkeiten

-0.45

parsedMessage

-0.43

género

-0.43

yntaxException

-0.42

 propOrder

-0.41

govina

-0.41

 ſche

-0.40

POSITIVE LOGITS

 slightly

0.44

 muffled

0.40

 mildly

0.39

 Slightly

0.39

闷

0.38

kwür

0.38

coltura

0.38

 Pauly

0.38

 lukewarm

0.38

 Pringle

0.37

Activations Density 0.378%

evidential phrases indicating contrasts or comparisons

contrasts or comparisons

No Comments

No Known Activations