INDEX

Explanations

the word "fire" or variants of it

oai_token-act-pair · gpt-3.5-turbo

instances of the word "fire."

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

ļé

-0.82

ĸļ

-0.82

 carbohyd

-0.73

achusetts

-0.71

 cellul

-0.70

Īè

-0.69

romeda

-0.69

 nont

-0.68

srf

-0.68

ockets

-0.67

POSITIVE LOGITS

nces

1.00

cia

0.90

lli

0.86

ly

0.86

lessly

0.83

ttes

0.81

 Dame

0.79

zza

0.79

les

0.79

tto

0.78

Activations Density 0.014%

the word "fire" or variants of it

instances of the word "fire."

No Comments

No Known Activations