INDEX

Explanations

references to programming interfaces or structures in code-related contexts

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

dialog interface programming

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

存于互联网档案馆

-0.56

fillType

-0.55

 compens

-0.48

bif

-0.47

 invis

-0.47

ssi

-0.46

 Invis

-0.46

Kis

-0.45

 Heil

-0.44

 bary

-0.44

POSITIVE LOGITS

 FormGroup

3.03

FormGroup

1.72

 FormControl

0.76

 ModelExpression

0.64

 FormBuilder

0.63

 étoient

0.59

 desmotivaciones

0.57

 Forderung

0.56

 gruesa

0.54

 GenerationType

0.53

Activations Density 0.000%

references to programming interfaces or structures in code-related contexts

dialog interface programming

No Comments

No Known Activations