INDEX

Explanations

names of specific locations or individuals

oai_token-act-pair · gpt-3.5-turbo

prominent names and locations associated with specific individuals or organizations

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

asury

-0.79

neau

-0.74

iller

-0.73

avier

-0.72

dimensional

-0.72

uren

-0.72

ilage

-0.71

arde

-0.70

apsed

-0.69

eared

-0.69

POSITIVE LOGITS

SHIP

0.79

 Tablet

0.71

Kop

0.71

é¾įå¥ĳå£«

0.69

 sidx

0.64

 Origins

0.62

 Scandinavian

0.61

 Broadcast

0.60

 Scandinav

0.59

 Munich

0.59

Activations Density 0.054%

names of specific locations or individuals

prominent names and locations associated with specific individuals or organizations

No Comments

No Known Activations