INDEX

Explanations

the beginning of sections or paragraphs in a document

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

research document intros

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 16-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_16/width_65k/average_l0_128

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.16.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

BatchNorm

-0.53

 tombé

-0.49

 всегда

-0.49

 détru

-0.47

 дописавши

-0.47

putstr

-0.47

 alltid

-0.46

 always

-0.45

 AssemblyCulture

-0.45

 pourtant

-0.44

POSITIVE LOGITS

 objectives

0.68

 scoping

0.68

 subpo

0.67

 scope

0.66

 output

0.66

SCOPE

0.64

 seconded

0.63

 deliverables

0.63

 subpoena

0.63

SharedDtor

0.63

Activations Density 0.079%

the beginning of sections or paragraphs in a document

research document intros

No Comments

No Known Activations