INDEX

Explanations

phrases related to news headlines

oai_token-act-pair · gpt-3.5-turbo

conjunctions and the use of the word "and."

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

IUM

-0.77

ysis

-0.76

renheit

-0.72

hower

-0.72

enegger

-0.71

beit

-0.71

IDA

-0.70

stocks

-0.68

ONSORED

-0.65

blance

-0.65

POSITIVE LOGITS

 Clyde

0.87

 mortar

0.74

mort

0.73

 nephew

0.68

 Camb

0.66

 Sisters

0.66

 Gall

0.65

 Dund

0.64

 Sutherland

0.64

 Koch

0.63

Activations Density 0.073%

phrases related to news headlines

conjunctions and the use of the word "and."

No Comments

No Known Activations