INDEX

Explanations

proper nouns, potentially related to locations or organizations

oai_token-act-pair · gpt-3.5-turbo

references to certain proper nouns or specific identifiers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

•Starts with {letter} features

No Comments

Negative Logits

arest

-0.79

arent

-0.72

ouse

-0.71

obyl

-0.67

agate

-0.67

ishops

-0.67

arium

-0.66

 Arlington

-0.66

ouses

-0.65

aving

-0.64

POSITIVE LOGITS

2.24

2.19

1.81

1.75

JD

1.69

JA

1.61

Je

1.59

JM

1.57

JC

1.55

Js

1.54

Activations Density 0.532%

proper nouns, potentially related to locations or organizations

references to certain proper nouns or specific identifiers

No Comments

No Known Activations