INDEX

Explanations

mentions of the city "Louisville"

oai_token-act-pair · gpt-3.5-turbo

mentions of the city of Louisville

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

BOOK

-0.77

policy

-0.71

vana

-0.70

-0.69

PsyNetMessage

-0.69

ãĤ¯

-0.69

ãģ¦

-0.67

vor

-0.67

ird

-0.66

opic

-0.65

POSITIVE LOGITS

 Slug

1.08

 Cardinals

0.90

 Courier

0.89

 Louisville

0.87

 Metropolitan

0.77

 Parish

0.76

iana

0.71

 Cable

0.71

 Pride

0.69

halla

0.69

Activations Density 0.007%

mentions of the city "Louisville"

mentions of the city of Louisville

No Comments

No Known Activations