INDEX

Explanations

specific actions or behaviors described in a narrative context

oai_token-act-pair · gpt-3.5-turbo

phrases indicating responses and inquiries in conversations

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

agate

-0.77

road

-0.75

isode

-0.75

zhou

-0.72

joice

-0.70

rift

-0.67

angular

-0.67

hazard

-0.65

hill

-0.64

hab

-0.63

POSITIVE LOGITS

 himself

1.34

his

1.03

 detractors

0.97

 opponents

0.94

 reporters

0.92

 others

0.91

 anyone

0.90

 everyone

0.90

 anybody

0.89

 teammates

0.89

Activations Density 0.996%

specific actions or behaviors described in a narrative context

phrases indicating responses and inquiries in conversations

No Comments

No Known Activations