INDEX

Explanations

references to notable individuals and their achievements

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

preceding nouns/pronouns

np_acts-logits-general · gemini-2.0-flash

unusual or specific details

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k

Configuration

google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

enumii

-0.66

psack

-0.66

 loisirs

-0.64

 algemene

-0.62

enumi

-0.62

 conséquence

-0.62

 kväll

-0.60

 nhàng

-0.60

 geweest

-0.59

 quaisquer

-0.58

POSITIVE LOGITS

 giant

0.76

 underwater

0.75

 tattooed

0.74

 Guinness

0.70

underwater

0.66

 robot

0.66

竟然

0.66

 toilet

0.66

 weirdly

0.65

 Hitler

0.64

Activations Density 0.559%

references to notable individuals and their achievements

preceding nouns/pronouns

unusual or specific details

No Comments

No Known Activations