INDEX

Explanations

the word "tone" at varying levels of intensity

oai_token-act-pair · gpt-3.5-turbo

references to tone and its various contextual implications

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

minster

-0.82

alus

-0.77

iott

-0.70

icipated

-0.70

essions

-0.70

arus

-0.68

riks

-0.68

joining

-0.66

anse

-0.66

actus

-0.66

POSITIVE LOGITS

 deaf

1.06

 tone

0.90

 Tone

0.85

 assurance

0.76

 dial

0.76

 shif

0.75

 tune

0.73

 tones

0.72

 amplification

0.65

xual

0.65

Activations Density 0.045%

the word "tone" at varying levels of intensity

references to tone and its various contextual implications

No Comments

No Known Activations