INDEX

Explanations

words related to manipulation or coercion

oai_token-act-pair · gpt-3.5-turbo

variations of the word "entertainment" or related terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 Responsibility

-0.78

 Spears

-0.76

å§«

-0.75

 Accountability

-0.70

 Jenner

-0.70

BILITIES

-0.69

 STAR

-0.68

pmwiki

-0.68

Universal

-0.66

å£«

-0.65

POSITIVE LOGITS

ailed

1.06

ourage

1.00

rave

1.00

renched

0.99

inence

0.96

ailing

0.95

uring

0.95

ent

0.94

rust

0.92

rench

0.91

Activations Density 0.007%

words related to manipulation or coercion

variations of the word "entertainment" or related terms

No Comments

No Known Activations