INDEX

Explanations

phrases where something/someone becomes strongly associated with something else

oai_token-act-pair · gpt-3.5-turbo

phrases related to transformation and becoming something new

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 balances

-0.68

 bene

-0.67

esides

-0.64

ups

-0.63

 survived

-0.60

 breaths

-0.59

 bribe

-0.56

 invented

-0.56

arton

-0.55

nab

-0.55

POSITIVE LOGITS

 fodder

0.78

alde

0.77

sonian

0.71

ZE

0.71

£

0.71

 anew

0.71

Ire

0.69

 scapego

0.69

iru

0.69

irl

0.68

Activations Density 0.203%

phrases where something/someone becomes strongly associated with something else

phrases related to transformation and becoming something new

No Comments

No Known Activations