INDEX

Explanations

negations or expressions of denial

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

not instead of

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.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

 houſe

-0.58

VIAF

-0.51

 purpoſe

-0.47

 cliquez

-0.44

httphttps

-0.44

igshid

-0.44

 GenerationType

-0.43

gól

-0.43

urtstag

-0.42

POSITIVE LOGITS

而非

0.64

而不是

0.60

 yalnızca

0.58

 malah

0.53

 sadece

0.52

Только

0.52

 nejen

0.51

 μόνο

0.51

not

0.50

 rather

0.50

Activations Density 0.024%

negations or expressions of denial

not instead of

No Comments

No Known Activations