INDEX

Explanations

expressions related to evaluating options and trade-offs

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

whether, considering, choices

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

GOTREF

-0.53

ۣ

-0.52

GIVEREF

-0.50

Derbyniad

-0.48

 चीज़ों

-0.47

BeginInit

-0.46

cèse

-0.45

preuve

-0.45

guisement

-0.45

 ſtate

-0.44

POSITIVE LOGITS

VersionUID

0.42

Bow

0.37

ScopeManager

0.36

 tempted

0.35

 Candidates

0.34

Considering

0.34

SF

0.34

bow

0.33

 Guille

0.33

 Trying

0.32

Activations Density 0.030%

expressions related to evaluating options and trade-offs

whether, considering, choices

No Comments

No Known Activations