INDEX

Explanations

mentions of the ACLU (American Civil Liberties Union)

oai_token-act-pair · gpt-3.5-turbo

mentions of civil rights organizations, particularly the ACLU

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

uled

-0.78

 Camb

-0.73

lass

-0.70

wagen

-0.70

prime

-0.69

dating

-0.68

olls

-0.67

erm

-0.66

rule

-0.65

hy

-0.65

POSITIVE LOGITS

 ACLU

1.06

 GOODMAN

0.86

 NAACP

0.86

 Fein

0.81

 Orchestra

0.80

 Liberties

0.74

 Consortium

0.72

 attorney

0.69

 challeng

0.67

ipation

0.66

Activations Density 0.006%

mentions of the ACLU (American Civil Liberties Union)

mentions of civil rights organizations, particularly the ACLU

No Comments

No Known Activations