INDEX

Explanations

people becoming interested or involved in various topics or activities

oai_token-act-pair · gpt-3.5-turbo

terms related to interest and significance 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 @ 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

ingham

-0.62

esides

-0.58

Bi

-0.58

oken

-0.57

 bene

-0.57

 Barnes

-0.57

 balances

-0.54

phia

-0.54

ongyang

-0.54

 Borderlands

-0.53

POSITIVE LOGITS

ãĤ¼

0.78

£

0.72

 anew

0.72

 References

0.70

 fodder

0.70

aez

0.70

 scapego

0.69

å§«

0.65

iru

0.65

/+

0.63

Activations Density 0.169%

people becoming interested or involved in various topics or activities

terms related to interest and significance in various contexts

No Comments

No Known Activations