INDEX

Explanations

scientific terms and experimental methods

oai_token-act-pair · gpt-3.5-turbo

elements related to detailed descriptions and conditions in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

jri

-0.69

 Vaugh

-0.63

âĢ¦]

-0.62

 Nare

-0.62

...]

-0.60

 Cannes

-0.60

 quota

-0.59

 Noir

-0.58

 Laksh

-0.57

 Malone

-0.54

POSITIVE LOGITS

oglu

0.72

mosp

0.68

 largeDownload

0.59

olin

0.56

 Minotaur

0.56

amina

0.55

BILITIES

0.54

oli

0.54

ctive

0.53

 Thermal

0.53

Activations Density 0.237%

scientific terms and experimental methods

elements related to detailed descriptions and conditions in various contexts

No Comments

No Known Activations