INDEX

Explanations

the word "un"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "un."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

URES

-0.70

RD

-0.70

wolves

-0.68

Home

-0.68

Series

-0.67

Previously

-0.66

OSH

-0.64

INGS

-0.63

Current

-0.62

Previous

-0.61

POSITIVE LOGITS

iver

0.96

 conflic

0.83

0.78

ter

0.73

 mund

0.73

pu

0.72

 pione

0.71

itud

0.70

nu

0.69

itaire

0.69

Activations Density 0.020%

the word "un"

occurrences of the word "un."

No Comments

No Known Activations