INDEX

Explanations

statements or phrases suggesting a strong point of view or assertion

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate expression or significance of a situation or opinion

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

untarily

-0.70

ammers

-0.68

IENCE

-0.66

swick

-0.66

iencies

-0.65

nces

-0.64

imentary

-0.64

cyclopedia

-0.63

href

-0.63

edia

-0.63

POSITIVE LOGITS

 strut

0.65

 Piper

0.63

Pry

0.62

CRC

0.59

 Rouge

0.59

 promising

0.59

 predicting

0.59

 marqu

0.54

 utopian

0.53

 doom

0.52

Activations Density 0.580%

statements or phrases suggesting a strong point of view or assertion

phrases that indicate expression or significance of a situation or opinion

No Comments

No Known Activations