INDEX

Explanations

references to Vice President Joe Biden

oai_token-act-pair · gpt-3.5-turbo

mentions of Joe Biden

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Ö¼

-0.75

ivities

-0.74

 ILCS

-0.74

Reviewer

-0.72

 Anarchy

-0.70

istance

-0.69

ortmund

-0.69

phis

-0.67

ivated

-0.67

ateur

-0.66

POSITIVE LOGITS

 Biden

1.25

zag

0.81

 Caucus

0.80

mire

0.77

 aide

0.76

jug

0.76

 appoint

0.74

 Jinping

0.74

 pard

0.72

bj

0.71

Activations Density 0.009%

references to Vice President Joe Biden

mentions of Joe Biden

No Comments

No Known Activations