INDEX

Explanations

verbs related to giving permission or enabling actions

oai_token-act-pair · gpt-3.5-turbo

references to users or participants in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

obia

-0.61

rection

-0.60

mbol

-0.56

leneck

-0.56

uncle

-0.55

 disappoint

-0.54

ULAR

-0.54

Sound

-0.53

 charm

-0.53

 Cipher

-0.52

POSITIVE LOGITS

 alike

1.01

who

0.97

hip

0.93

 wishing

0.86

 everywhere

0.86

hips

0.85

who

0.83

 worldwide

0.81

paces

0.79

 nationwide

0.78

Activations Density 0.131%

verbs related to giving permission or enabling actions

references to users or participants in various contexts

No Comments

No Known Activations