INDEX

Explanations

names related to people or places

oai_token-act-pair · gpt-3.5-turbo

names of individuals or entities, particularly those related to diet and health

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

iasm

-0.75

anamo

-0.74

orthy

-0.74

Ń·

-0.73

 eleph

-0.72

iris

-0.71

izoph

-0.71

orrow

-0.70

inx

-0.70

osate

-0.69

POSITIVE LOGITS

ership

0.88

rich

0.86

ering

0.83

Eck

0.82

 Harden

0.81

 Braun

0.78

 Diet

0.78

osterone

0.77

 Coke

0.76

 Kling

0.75

Activations Density 0.022%

names related to people or places

names of individuals or entities, particularly those related to diet and health

No Comments

No Known Activations