INDEX

Explanations

phrases that introduce a statement or topic

oai_token-act-pair · gpt-3.5-turbo

phrases that introduce or reference topics of discussion

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

~~~~~~~~

-0.70

eries

-0.69

PIN

-0.69

î

-0.68

knit

-0.67

CTV

-0.67

rils

-0.67

ï¿½ï¿½

-0.66

\'

-0.65

frac

-0.65

POSITIVE LOGITS

 specifics

0.79

 concluding

0.78

 naming

0.76

 praising

0.74

 mentioning

0.71

 dismissing

0.70

 parting

0.69

 correcting

0.69

 criticism

0.68

ebus

0.67

Activations Density 0.149%

phrases that introduce a statement or topic

phrases that introduce or reference topics of discussion

No Comments

No Known Activations