INDEX

Explanations

coding language constructs

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

AI schedules and 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

ilions

-0.40

BeginContext

-0.40

 myö

-0.36

ThroughAttribute

-0.36

 descended

-0.35

 galore

-0.34

 explí

-0.34

 jsonify

-0.33

 ostavi

-0.33

sprü

-0.33

POSITIVE LOGITS

GEBURTSDATUM

0.64

Panamoan

0.57

 Administrativna

0.55

 """",

0.52

vac

0.50

 <=",

0.50

出版年

0.49

TagMode

0.49

 invokingState

0.47

UseVisualStyle

0.47

Activations Density 0.017%

coding language constructs

AI schedules and choices

No Comments

No Known Activations