INDEX

Explanations

phrases indicating user actions and decision-making

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 7-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_7/width_16k/average_l0_69

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.7.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

']],

-0.74

 isComment

-0.69

pexpr

-0.68

twimg

-0.66

 оригіналу

-0.61

autaire

-0.60

 pylint

-0.60

theless

-0.59

[::-

-0.59

AndEndTag

-0.58

POSITIVE LOGITS

 propOrder

0.68

 ExecuteAsync

0.63

 مشين

0.55

SerializeField

0.50

이다

0.48

else

0.47

 Chances

0.46

 else

0.46

 endif

0.46

 ujednoznacz

0.45

Activations Density 0.176%

phrases indicating user actions and decision-making

No Comments

No Known Activations