INDEX

Explanations

mentions of restaurants

oai_token-act-pair · gpt-3.5-turbo

mentions of restaurants

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

ioxide

-0.77

 Matter

-0.74

 Measures

-0.68

MAT

-0.66

 Latest

-0.66

 Weld

-0.64

OX

-0.63

 Belarus

-0.63

DPR

-0.62

 Powers

-0.62

POSITIVE LOGITS

 Restaur

1.08

 restaurants

1.05

 eater

1.05

 restaurant

1.00

eteria

0.96

 Restaurant

0.90

isine

0.90

ateurs

0.89

izza

0.89

 establishments

0.88

Activations Density 0.019%

mentions of restaurants

mentions of restaurants

No Comments

No Known Activations