INDEX

Explanations

phrases that indicate legal or procedural contexts

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

relation or comparison

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.31.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

abestanden

-0.37

 Featured

-0.33

 уста

-0.32

intahan

-0.31

 Welcome

-0.30

IVEREF

-0.30

 виправивши

-0.30

mány

-0.30

oyn

-0.29

 yerine

-0.28

POSITIVE LOGITS

ponses

0.68

posób

0.66

 perbedaan

0.66

 możliwość

0.66

consistency

0.65

 možnost

0.65

ilarities

0.64

 differences

0.63

 możliwości

0.61

izability

0.60

Activations Density 0.499%

phrases that indicate legal or procedural contexts

relation or comparison

No Comments

No Known Activations