INDEX

Explanations

references to initializations and initial states in code contexts

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

initialization related to code

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

New Auto-Interp

Top Features by Cosine Similarity

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

Configuration

google/gemma-scope-9b-it-res/layer_9/width_131k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.9.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

 bVar

-0.48

 vote

-0.46

föra

-0.43

 onError

-0.42

 sendMessage

-0.41

onStop

-0.41

fört

-0.40

CODEC

-0.40

 varandra

-0.40

 removeFrom

-0.39

POSITIVE LOGITS

initial

1.63

Initial

1.58

 Initial

1.44

INITIAL

1.43

 initial

1.35

setInitial

1.34

 INITIAL

1.29

Intial

1.13

 initiale

1.03

 inicial

0.98

Activations Density 0.009%

references to initializations and initial states in code contexts

initialization related to code

No Comments

No Known Activations