INDEX

Explanations

specific locations or geographical features

oai_token-act-pair · gpt-3.5-turbo

terms related to geographical features and settings

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

iak

-0.93

kus

-0.91

rations

-0.84

lear

-0.83

vals

-0.81

ration

-0.78

ļ

-0.77

oway

-0.76

este

-0.75

orie

-0.74

POSITIVE LOGITS

APR

0.62

 whilst

0.61

xual

0.60

 overlook

0.59

 AFTER

0.55

 surrounded

0.55

 READ

0.55

 EDITION

0.54

 fuss

0.54

 hitch

0.54

Activations Density 0.445%

specific locations or geographical features

terms related to geographical features and settings

No Comments

No Known Activations