INDEX

Explanations

negations or statements that express doubt or disagreement in a context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 15-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_15/width_16k/average_l0_78

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.15.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

ead

-0.46

-0.45

र्व

-0.45

due

-0.44

Due

-0.42

Due

-0.42

 proper

-0.42

Proper

-0.40

Vector

-0.40

LEADING

-0.40

POSITIVE LOGITS

 necessarily

0.93

 necessariamente

0.93

 necesariamente

0.85

 quelconque

0.80

necessarily

0.75

 neceff

0.73

 nécessairement

0.73

المكان

0.73

 leſs

0.70

oredCriteria

0.69

Activations Density 0.455%

negations or statements that express doubt or disagreement in a context

No Comments

No Known Activations