INDEX

Explanations

references to the prefix "re-" indicating repetition or retrieval

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

Tokens followed by "re" (prefix)

np_acts-logits-general · gemini-2.0-flash

re- followed by a verb

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

ftagPool

-0.83

 définiti

-0.81

 épu

-0.74

 chronique

-0.72

 détru

-0.72

 pédagog

-0.72

antMatchers

-0.71

 réguli

-0.68

 supplémentaires

-0.67

 feroit

-0.67

POSITIVE LOGITS

re

0.71

 насељу

0.56

StoryboardSegue

0.52

ANSA

0.51

rej

0.50

0.49

makeText

0.49

res

0.47

AccessFile

0.47

 launch

0.45

Activations Density 0.104%

references to the prefix "re-" indicating repetition or retrieval

Tokens followed by "re" (prefix)

re- followed by a verb

No Comments

No Known Activations