INDEX

Explanations

information sources or references being cited or mentioned

oai_token-act-pair · gpt-3.5-turbo

references to sources of information or evidence

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

Ħ¢

-0.99

aeper

-0.77

eer

-0.68

 Potato

-0.65

eers

-0.65

ternity

-0.63

ipop

-0.61

 Nept

-0.61

ecast

-0.60

«ĺ

-0.60

POSITIVE LOGITS

 sources

1.45

 Sources

1.37

ources

1.21

Sources

1.08

source

1.07

 source

1.05

 sourcing

0.97

ource

0.85

etter

0.80

 Source

0.79

Activations Density 0.016%

information sources or references being cited or mentioned

references to sources of information or evidence

No Comments

No Known Activations