INDEX

Explanations

occurrences of parentheses and phrases or words related to questioning or clarification

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 4-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_4/width_16k/average_l0_124

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.4.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

-0.71

omiast

-0.69

IBAction

-0.65

RenderAtEndOf

-0.64

 함께

-0.57

érial

-0.57

Переваги

-0.57

"]}

-0.56

Cordialement

-0.56

 CURIAM

-0.56

POSITIVE LOGITS

but

0.63

baomidou

0.62

gulier

0.59

 although

0.59

 thankfully

0.57

Rè

0.56

er

0.56

 luckily

0.55

almost

0.55

 unfortunately

0.55

Activations Density 0.223%

occurrences of parentheses and phrases or words related to questioning or clarification

No Comments

No Known Activations