INDEX

Explanations

names of individuals

oai_token-act-pair · gpt-3.5-turbo

references to individual names and key figures in the context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

usterity

-0.74

onymous

-0.70

 Chilean

-0.70

vironment

-0.70

ructure

-0.70

ulhu

-0.69

ainment

-0.67

eral

-0.66

 conspicuous

-0.64

alid

-0.63

POSITIVE LOGITS

 Jacobs

1.31

tones

0.90

aunders

0.78

mann

0.68

 Amon

0.67

inelli

0.67

 intrins

0.67

robe

0.66

aign

0.66

hetti

0.66

Activations Density 0.004%

names of individuals

references to individual names and key figures in the context

No Comments

No Known Activations