INDEX

Explanations

terms related to written communication, specifically memos

oai_token-act-pair · gpt-3.5-turbo

instances of the word "memo."

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

reek

-0.67

Interstitial

-0.66

Fen

-0.66

orks

-0.63

tool

-0.63

Pyr

-0.63

rations

-0.62

rises

-0.62

 Democr

-0.62

respect

-0.61

POSITIVE LOGITS

 memo

1.19

andum

0.97

 memos

0.90

ariat

0.86

ufact

0.84

 synopsis

0.81

ographed

0.79

emort

0.78

osal

0.78

isode

0.78

Activations Density 0.013%

terms related to written communication, specifically memos

instances of the word "memo."

No Comments

No Known Activations