INDEX

Explanations

specific names and terms related to various individuals, locations, and organizations

oai_token-act-pair · gpt-3.5-turbo

proper names, particularly surnames and their frequency in the text

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

spring

-0.89

ivals

-0.82

ename

-0.79

raints

-0.75

 Crusader

-0.74

inals

-0.71

 Columb

-0.71

 Scotch

-0.69

Lex

-0.69

central

-0.68

POSITIVE LOGITS

 Wheeler

1.25

Wheel

1.04

 mosqu

0.90

\\\\\\\\

0.86

wheel

0.80

leigh

0.76

 shenan

0.75

worms

0.75

oshenko

0.72

 dismant

0.72

Activations Density 0.007%

specific names and terms related to various individuals, locations, and organizations

proper names, particularly surnames and their frequency in the text

No Comments

No Known Activations