INDEX

Explanations

mentions of cafes

oai_token-act-pair · gpt-3.5-turbo

mentions of cafes

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

PATH

-0.70

iod

-0.69

sburg

-0.65

methyl

-0.62

RAY

-0.62

ansom

-0.61

kus

-0.60

INGS

-0.58

ologically

-0.58

female

-0.58

POSITIVE LOGITS

 cafe

1.16

 cafes

1.14

 cafÃ©

1.05

 Cafe

0.99

caf

0.97

 racer

0.90

eteria

0.89

 CafÃ©

0.87

 kios

0.86

ecake

0.85

Activations Density 0.009%

mentions of cafes

mentions of cafes

No Comments

No Known Activations