INDEX

Explanations

phrases indicating refusals or lack of information

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

Tokens preceding "not"

np_acts-logits-general · gemini-2.0-flash

absence of comment

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.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

styleType

-0.57

TagHelper

-0.56

 politely

-0.52

 respe

-0.50

Успе

-0.50

 posib

-0.50

 trusting

-0.50

たまたま

-0.50

correctly

-0.49

PerformLayout

-0.49

POSITIVE LOGITS

 comment

0.82

 commenting

0.73

 Comment

0.73

comment

0.71

 specific

0.71

specific

0.67

Comment

0.64

 specifics

0.64

Specific

0.64

 spezif

0.64

Activations Density 0.495%

phrases indicating refusals or lack of information

Tokens preceding "not"

absence of comment

No Comments

No Known Activations