INDEX

Explanations

terms related to peer review processes

oai_token-act-pair · gpt-3.5-turbo

specific references to legal terminology or illicit activities

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

eton

-0.75

aline

-0.74

ylon

-0.70

selves

-0.69

Fly

-0.67

HAM

-0.67

 Avenger

-0.66

 Hang

-0.65

 Ends

-0.64

self

-0.63

POSITIVE LOGITS

utable

0.78

 crochet

0.72

 certify

0.67

cess

0.63

 artisan

0.62

utive

0.62

UGE

0.61

vou

0.59

batch

0.58

owicz

0.58

Activations Density 0.000%

terms related to peer review processes

specific references to legal terminology or illicit activities

No Comments

No Known Activations