INDEX

Explanations

emphasized references to specific subjects or objects in the text

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

setVerticalGroup

-0.60

RectangleBorder

-0.49

adaptiveStyles

-0.46

 AssemblyVersion

-0.43

idual

-0.42

 ✭✭

-0.42

 monter

-0.41

EXTERN

-0.40

 Mounted

-0.40

...@

-0.40

POSITIVE LOGITS

 means

0.57

Gön

0.52

 anſ

0.52

may

0.51

complexContent

0.51

 often

0.51

 ensures

0.49

 hieronder

0.48

 also

0.48

 asegur

0.48

Activations Density 0.009%

emphasized references to specific subjects or objects in the text

No Comments

No Known Activations