INDEX

Explanations

references to the city of Birmingham

oai_token-act-pair · gpt-3.5-turbo

references to the city of Birmingham

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

afort

-0.85

ivot

-0.79

sth

-0.73

oward

-0.72

ential

-0.70

cess

-0.70

SHIP

-0.69

UAL

-0.69

entially

-0.68

uador

-0.67

POSITIVE LOGITS

mingham

1.17

 accents

0.81

mouth

0.80

shire

0.80

keley

0.79

thouse

0.77

 Blitz

0.76

©¶æ¥µ

0.71

sburg

0.71

 Yard

0.70

Activations Density 0.007%

references to the city of Birmingham

references to the city of Birmingham

No Comments

No Known Activations