INDEX

Explanations

mentions of political figures holding positions of power

oai_token-act-pair · gpt-3.5-turbo

mentions of "governor" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

clips

-0.64

formance

-0.61

iths

-0.61

rences

-0.60

tle

-0.60

asons

-0.60

occ

-0.59

kies

-0.58

ATL

-0.58

Ent

-0.58

POSITIVE LOGITS

 governor

3.95

 Governor

2.95

 governors

2.59

 mayor

2.11

Gov

2.05

 gubernatorial

2.00

Govern

1.95

Gov

1.80

 Governors

1.71

govern

1.63

Activations Density 0.014%

mentions of political figures holding positions of power

mentions of "governor" in various contexts

No Comments

No Known Activations