INDEX

Explanations

references to specific names or entities that include the word "Ble", with varying degrees of intensity

oai_token-act-pair · gpt-3.5-turbo

mentions of the publication "Bleacher Report" and related terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Helsinki

-0.78

heid

-0.74

ahime

-0.74

 perjury

-0.72

ional

-0.72

orney

-0.71

afort

-0.69

omez

-0.68

å§«

-0.67

 Pathfinder

-0.67

POSITIVE LOGITS

bles

0.83

acher

0.81

pedia

0.81

zz

0.79

¸

0.79

 Digest

0.79

erd

0.78

keye

0.78

eding

0.76

eker

0.76

Activations Density 0.005%

references to specific names or entities that include the word "Ble", with varying degrees of intensity

mentions of the publication "Bleacher Report" and related terms

No Comments

No Known Activations