INDEX

Explanations

recipes or instructions

oai_token-act-pair · gpt-3.5-turbo

words related to appreciation and recognition

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

haz

-0.73

 Bermuda

-0.68

 Catalyst

-0.68

 boarding

-0.66

 Belt

-0.66

 Hallow

-0.65

 Butt

-0.65

 Canary

-0.63

 purse

-0.63

 Span

-0.63

POSITIVE LOGITS

reci

1.42

ocity

1.10

ating

0.98

apy

0.97

issance

0.96

ated

0.96

bled

0.93

ate

0.93

ased

0.92

asing

0.90

Activations Density 0.015%

recipes or instructions

words related to appreciation and recognition

No Comments

No Known Activations