INDEX

Explanations

terms related to functional capabilities or activities

oai_token-act-pair · gpt-3.5-turbo

references to functional attributes or concepts

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

 Twain

-0.78

Swe

-0.73

Iss

-0.70

 Colleges

-0.69

yang

-0.69

 Preferred

-0.69

 Lank

-0.68

Brow

-0.68

 Mothers

-0.67

Roy

-0.67

POSITIVE LOGITS

ities

0.99

istic

0.97

tarian

0.96

acies

0.93

ative

0.92

tarians

0.91

iencies

0.88

eatures

0.87

functional

0.87

atically

0.87

Activations Density 0.007%

terms related to functional capabilities or activities

references to functional attributes or concepts

No Comments

No Known Activations