INDEX

Explanations

conditional phrases and logical connections in programming contexts

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

Spanish and French conjunctions and questions

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

BeginContext

-0.69

 חיצוניים

-0.64

SequentialGroup

-0.61

 CURIAM

-0.59

IntoConstraints

-0.59

Derbyniad

-0.58

RenderAtEndOf

-0.57

-0.56

tagHelperRunner

-0.55

 '\\;'

-0.55

POSITIVE LOGITS

is

0.67

has

0.64

 have

0.62

can

0.59

 actually

0.52

are

0.52

 feels

0.47

 takes

0.46

se

0.46

 holds

0.45

Activations Density 0.011%

conditional phrases and logical connections in programming contexts

Spanish and French conjunctions and questions

No Comments

No Known Activations