INDEX

Explanations

references to specific locations or entities

oai_token-act-pair · gpt-3.5-turbo

specific numerical identifiers or references related to games, places, and notable individuals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

rh

-0.59

enance

-0.57

raints

-0.56

orf

-0.55

upuncture

-0.53

wic

-0.53

ctr

-0.52

udic

-0.52

 istg

-0.52

ded

-0.51

POSITIVE LOGITS

 advoc

0.81

 corrid

0.80

 Citiz

0.76

 cryst

0.75

 streng

0.71

 unden

0.69

 contrace

0.68

 princ

0.67

 seiz

0.66

 disadvant

0.66

Activations Density 3.945%

references to specific locations or entities

specific numerical identifiers or references related to games, places, and notable individuals

No Comments

No Known Activations