INDEX

Explanations

terms related to endorsements and recommendations

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

endorsement and acceptance

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.31.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



-0.56

PyExc

-0.52

 ligiloj

-0.48

bleau

-0.46

bridi

-0.44

ismet

-0.44

 sebou

-0.44

 yourselves

-0.43

 spits

-0.42

 penalties

-0.42

POSITIVE LOGITS

 endorsement

0.42

 endorse

0.40

 Endor

0.40

 AssemblyCompany

0.40

EnableWeb

0.39

AndEndTag

0.39

 Endless

0.38

endor

0.35

 nation

0.35

InjectAttribute

0.35

Activations Density 0.000%

terms related to endorsements and recommendations

endorsement and acceptance

No Comments

No Known Activations