INDEX

Explanations

phrases related to technical specifications or structured data

oai_token-act-pair · gpt-3.5-turbo

numerical data and references to structured information

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

heit

-0.76

 redes

-0.66

cius

-0.66

anchester

-0.65

 Gutierrez

-0.62

 nomine

-0.62

 Medina

-0.59

 Kath

-0.58

 showc

-0.58

cest

-0.58

POSITIVE LOGITS

fif

0.84

fig

0.79

Ïī

0.78

oshenko

0.77

scl

0.75

 è£ıè

0.72

drm

0.71

abytes

0.69

ohm

0.68

undred

0.67

Activations Density 0.219%

phrases related to technical specifications or structured data

numerical data and references to structured information

No Comments

No Known Activations