INDEX

Explanations

phrases related to the reliability or integrity of something

oai_token-act-pair · gpt-3.5-turbo

phrases indicating reliability and assessment

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

rev

-0.78

utan

-0.73

ESA

-0.71

KK

-0.70

monton

-0.69

HD

-0.67

 largeDownload

-0.66

rer

-0.66

rir

-0.66

orthy

-0.65

POSITIVE LOGITS

 these

0.74

 Nanto

0.74

 sorts

0.70

 warfare

0.65

 humankind

0.64

 those

0.63

emale

0.61

 mankind

0.61

our

0.60

 storytelling

0.59

Activations Density 0.174%

phrases related to the reliability or integrity of something

phrases indicating reliability and assessment

No Comments

No Known Activations