INDEX

Explanations

names of locations and organizations

oai_token-act-pair · gpt-3.5-turbo

terms and references related to a specific location or street name

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

ascript

-0.79

ilial

-0.77

catentry

-0.75

 Qiao

-0.70

theless

-0.68

redo

-0.64

Versions

-0.62

 mantle

-0.59

arial

-0.59

phony

-0.59

POSITIVE LOGITS

 Lumpur

0.96

EStream

0.86

EStreamFrame

0.86

ikuman

0.75

bye

0.74

wagen

0.72

nik

0.71

utra

0.70

boxing

0.69

nis

0.66

Activations Density 0.560%

names of locations and organizations

terms and references related to a specific location or street name

No Comments

No Known Activations