INDEX

Explanations

sequences of letters or symbols related to genetic information or biology

oai_token-act-pair · gpt-3.5-turbo

genetic terms and molecular identifiers

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ãĥ¼ãĥĨãĤ£

-0.84

 Tradable

-0.83

terday

-0.82

advertising

-0.79

ãĥĩãĤ£

-0.79

ãĥĥ

-0.76

éĹĺ

-0.76

official

-0.75

Privacy

-0.71

biz

-0.70

POSITIVE LOGITS

 concentrations

1.04

 Î²

1.02

 genes

1.00

 lesions

0.98

 inhibition

0.98

Î²

0.98

Î±

0.96

 bacterial

0.96

 allele

0.96

 induced

0.96

Activations Density 0.226%

sequences of letters or symbols related to genetic information or biology

genetic terms and molecular identifiers

No Comments

No Known Activations