INDEX

Explanations

terms related to recognition or acknowledgment of achievement or impact

oai_token-act-pair · gpt-3.5-turbo

expressions and terminology related to impact and influential works

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

creen

-0.86

ãģ®éŃĶ

-0.82

halla

-0.76

NetMessage

-0.73

çĦ

-0.73

anguage

-0.71

TERN

-0.71

Downloadha

-0.70

lees

-0.69

RAFT

-0.68

POSITIVE LOGITS

ressive

1.05

ulse

1.03

acts

1.03

ression

0.96

aired

0.94

actor

0.90

iral

0.88

acet

0.88

Imp

0.87

ropri

0.86

Activations Density 0.009%

terms related to recognition or acknowledgment of achievement or impact

expressions and terminology related to impact and influential works

No Comments

No Known Activations