INDEX

Explanations

legislative terms and vocabulary

oai_token-act-pair · gpt-3.5-turbo

repeated mentions of legislation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

erd

-0.89

lighting

-0.76

elsen

-0.69

 Canal

-0.67

 Mous

-0.64

Sul

-0.64

 sailor

-0.63

ija

-0.62

raction

-0.62

 Forsaken

-0.62

POSITIVE LOGITS

 enacted

1.01

arians

0.92

iard

0.90

 enforcement

0.88

 authorizing

0.86

ĸļ

0.84

ional

0.84

clinton

0.81

 passed

0.80

 legalizing

0.79

Activations Density 0.014%

legislative terms and vocabulary

repeated mentions of legislation

No Comments

No Known Activations