INDEX

Explanations

numerical values or numerical references (e.g. years, numbers)

oai_token-act-pair · gpt-3.5-turbo

numerical values and their 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

Synopsis

-0.57

 Canaver

-0.54

lishing

-0.53

Newsletter

-0.52

 ÂŃ

-0.51

Story

-0.50

widget

-0.49

 Narendra

-0.47

UTERS

-0.46

ÂŃ

-0.46

POSITIVE LOGITS

sic

0.56

 nomine

0.53

anwhile

0.50

udos

0.48

 compan

0.48

 natives

0.46

 latter

0.46

 comr

0.46

abo

0.45

 interchange

0.44

Activations Density 1.818%

numerical values or numerical references (e.g. years, numbers)

numerical values and their contexts

No Comments

No Known Activations