INDEX

Explanations

mentions of political figures, particularly lawmakers

oai_token-act-pair · gpt-3.5-turbo

references to lawmakers and legislative actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

ãĤ¨ãĥ«

-0.92

phal

-0.72

âĶĢâĶĢâĶĢâĶĢ

-0.72

omorph

-0.71

Tur

-0.68

istic

-0.66

ļé

-0.64

Customer

-0.63

 Royale

-0.63

 Palest

-0.62

POSITIVE LOGITS

hips

1.16

hip

0.99

arians

0.94

ervatives

0.89

woman

0.84

 unanimously

0.80

 voted

0.78

 committees

0.77

rieve

0.76

doms

0.76

Activations Density 0.040%

mentions of political figures, particularly lawmakers

references to lawmakers and legislative actions

No Comments

No Known Activations