INDEX

Explanations

arguments and claims related to evaluations and decisions

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

verbs followed by complements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 تضيفلها

-0.74

rungsseite

-0.57

 CreateTagHelper

-0.56

contentLoaded

-0.52

complexContent

-0.52

 gyhoeddwyd

-0.52

mybatisplus

-0.51

BibitemShut

-0.50

 otomatig

-0.50

 estekak

-0.49

POSITIVE LOGITS

UBER

0.42

imaan

0.37

agissait

0.36

ligen

0.35

klo

0.35

 Dign

0.35

ima

0.35

otrop

0.34

RepeatedField

0.34

igem

0.33

Activations Density 0.214%

arguments and claims related to evaluations and decisions

verbs followed by complements

No Comments

No Known Activations