INDEX

Explanations

mentions of corruption or corporate-related terms

oai_token-act-pair · gpt-3.5-turbo

references to "corporate" 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 @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

CRIP

-0.84

 Furious

-0.64

 unaccompanied

-0.64

icter

-0.63

 ACTION

-0.62

Jol

-0.61

ICE

-0.60

ELS

-0.60

 Hunters

-0.60

insert

-0.60

POSITIVE LOGITS

porate

1.73

respond

1.54

por

1.51

ruption

1.50

rection

1.43

relation

1.40

poral

1.40

ridor

1.36

rupt

1.34

rupted

1.34

Activations Density 0.021%

mentions of corruption or corporate-related terms

references to "corporate" and related terms

No Comments

No Known Activations