INDEX

Explanations

names of locations or places, especially with unique or distinctive spellings

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

catentry

-0.77

 fict

-0.62

 mercy

-0.62

Assembly

-0.60

 bene

-0.59

 Qiao

-0.59

cess

-0.59

 illeg

-0.58

advertisement

-0.58

UCT

-0.57

POSITIVE LOGITS

nen

1.13

 Lumpur

1.07

zie

0.96

ataka

0.93

inski

0.93

enei

0.92

EStream

0.91

sky

0.90

lyak

0.88

nik

0.87

Activations Density 0.269%

names of locations or places, especially with unique or distinctive spellings

No Comments

No Known Activations