INDEX

Explanations

numerical data presented in a structured format

oai_token-act-pair · gpt-3.5-turbo

the word "for" 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 @ 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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

chenko

-0.77

iHUD

-0.71

edin

-0.71

bleacher

-0.70

Raid

-0.68

é¾

-0.66

pkg

-0.66

ulous

-0.65

itably

-0.64

amed

-0.64

POSITIVE LOGITS

 purposes

1.32

bidden

1.06

 example

1.04

 sake

0.97

geries

0.97

 instance

0.95

aging

0.95

 comparison

0.91

cible

0.89

 periods

0.86

Activations Density 0.186%

numerical data presented in a structured format

the word "for" in various contexts

No Comments

No Known Activations