INDEX

Explanations

names of people or places

oai_token-act-pair · gpt-3.5-turbo

mentions of individuals or their names

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

 replay

-0.68

AFTA

-0.64

cffffcc

-0.63

 Proced

-0.62

milo

-0.62

 coma

-0.59

 Rats

-0.58

 blot

-0.58

 Solitaire

-0.58

 psychiat

-0.58

POSITIVE LOGITS

enthal

0.98

igans

0.93

enger

0.91

know

0.80

qt

0.80

hed

0.76

é¾įåĸļå£«

0.75

erd

0.75

ington

0.75

eller

0.74

Activations Density 0.067%

names of people or places

mentions of individuals or their names

No Comments

No Known Activations