INDEX

Explanations

expressions of uncertainty or questioning

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

 pouvoit

-0.50

matchCondition

-0.50

 konuş

-0.48

 informée

-0.44

postolic

-0.43

 AssemblyVersion

-0.43

 Nantes

-0.43

 auroit

-0.42

労

-0.42

 FANT

-0.41

POSITIVE LOGITS

 již

0.40

 null

0.37

 évidence

0.36

Alike

0.35

 predictability

0.35

 endfor

0.34

afs

0.34

 déjà

0.34

 joke

0.34

 familiarity

0.34

Activations Density 0.064%

expressions of uncertainty or questioning

No Comments

No Known Activations