INDEX

Explanations

mentions of the word "bronze"

oai_token-act-pair · gpt-3.5-turbo

references to bronze medals and awards

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

URA

-0.79

dfx

-0.72

ournal

-0.71

rina

-0.68

ruct

-0.66

ARP

-0.66

oci

-0.66

AMI

-0.66

ATA

-0.65

ANG

-0.65

POSITIVE LOGITS

 bronze

0.96

 plaque

0.92

 medal

0.90

 sculpture

0.89

 sculptures

0.84

fish

0.83

 statue

0.83

 statues

0.77

Age

0.76

icum

0.74

Activations Density 0.007%

mentions of the word "bronze"

references to bronze medals and awards

No Comments

No Known Activations