INDEX

Explanations

words related to clothing, specifically tops

oai_token-act-pair · gpt-3.5-turbo

references to "top" in various contexts

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

 Arri

-0.64

arij

-0.63

AUD

-0.62

 Gaul

-0.61

ufact

-0.60

riages

-0.59

 Hurricanes

-0.57

 warr

-0.57

igned

-0.57

[|

-0.56

POSITIVE LOGITS

top

1.14

TOP

1.12

most

1.11

bottom

1.04

mast

1.03

ographical

0.93

Top

0.91

ronics

0.89

deck

0.85

ICS

0.84

Activations Density 0.012%

words related to clothing, specifically tops

references to "top" in various contexts

No Comments

No Known Activations