INDEX

Explanations

phrases related to peptides or medical conditions involving peptides

oai_token-act-pair · gpt-3.5-turbo

references to peptides and related biochemical concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

é¾

-0.89

 GOODMAN

-0.73

schild

-0.70

 nets

-0.66

enegger

-0.66

 Kirin

-0.65

nets

-0.63

worthiness

-0.63

 swirling

-0.62

ainers

-0.61

POSITIVE LOGITS

anut

1.02

pe

1.01

oples

0.99

rer

0.98

cific

0.98

asant

0.96

pper

0.92

OTUS

0.92

ller

0.91

aks

0.91

Activations Density 0.010%

phrases related to peptides or medical conditions involving peptides

references to peptides and related biochemical concepts

No Comments

No Known Activations