INDEX

Explanations

references to structured data or attributes in a data schema

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

restructure data objects

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

 MainAxisSize

-0.73

UserScript

-0.69

 pérd

-0.67

TypedDataSet

-0.65

 мәкал

-0.64

oredCriteria

-0.63

pgterms

-0.63

thenReturn

-0.60

ientras

-0.59

setVerticalGroup

-0.59

POSITIVE LOGITS

<tr>

0.53

 ditto

0.43

 curiously

0.43

Church

0.42

Flo

0.42

 Lagi

0.42

בו

0.41

 Bind

0.41

Two

0.40

 paraly

0.40

Activations Density 0.005%

references to structured data or attributes in a data schema

restructure data objects

No Comments

No Known Activations