INDEX

Explanations

mentions of legal or bureaucratic entities

oai_token-act-pair · gpt-3.5-turbo

references to "entities" 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 @ 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

Progress

-0.71

edo

-0.68

stroke

-0.65

iPhone

-0.65

jin

-0.65

NAD

-0.62

joy

-0.62

afa

-0.62

geon

-0.60

 Prec

-0.60

POSITIVE LOGITS

 entity

1.05

 entities

0.92

chnology

0.90

 owned

0.76

rance

0.75

 owners

0.74

terness

0.74

 owner

0.73

 Entity

0.73

 accredited

0.71

Activations Density 0.013%

mentions of legal or bureaucratic entities

references to "entities" in various contexts

No Comments

No Known Activations