INDEX

Explanations

locations or situations

oai_token-act-pair · gpt-3.5-turbo

phrases expressing locations or states of being

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 equivalents

-0.78

 respectively

-0.71

orously

-0.62

letters

-0.60

 cheaply

-0.57

threat

-0.57

 threat

-0.56

 sidx

-0.55

 srfAttach

-0.55

alions

-0.55

POSITIVE LOGITS

 nutshell

0.67

----------------------------------------------------------------

0.67

 unfold

0.63

 behold

0.63

------------------------

0.63

ovie

0.55

Ax

0.55

laughs

0.55

Reviewer

0.54

Alz

0.54

Activations Density 0.631%

locations or situations

phrases expressing locations or states of being

No Comments

No Known Activations