INDEX

Explanations

legal or legislative terms

oai_token-act-pair · gpt-3.5-turbo

references to specific legal or regulatory provisions

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

Magikarp

-0.69

ampa

-0.65

ritz

-0.65

 Garfield

-0.62

 temperament

-0.62

psy

-0.61

 Doodle

-0.60

MAT

-0.60

hiro

-0.60

 scalp

-0.59

POSITIVE LOGITS

ally

1.04

ality

0.99

witz

0.96

ional

0.92

alities

0.88

arians

0.87

arian

0.85

ttes

0.80

mental

0.79

ions

0.76

Activations Density 0.019%

legal or legislative terms

references to specific legal or regulatory provisions

No Comments

No Known Activations