INDEX

Explanations

references to a specific location, Tasmania

oai_token-act-pair · gpt-3.5-turbo

mentions of specific geographic locations and key technical terms

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

itures

-0.68

theless

-0.63

 arrest

-0.62

edly

-0.60

 continents

-0.59

rising

-0.58

quez

-0.58

unknown

-0.58

ufact

-0.57

ski

-0.56

POSITIVE LOGITS

onomy

0.94

ribute

0.91

uple

0.88

uesday

0.87

olkien

0.87

olerance

0.85

onga

0.84

pole

0.82

 Gingrich

0.79

oxin

0.79

Activations Density 0.115%

references to a specific location, Tasmania

mentions of specific geographic locations and key technical terms

No Comments

No Known Activations