INDEX

Explanations

references to the name "nam" combined with specific surrounding context

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "nam" and possibly variations related to names or important figures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

 Progressive

-0.66

 Blackwell

-0.64

Introduced

-0.62

Premium

-0.62

 Subtle

-0.62

UID

-0.61

 record

-0.61

 cider

-0.60

ECB

-0.59

Leary

-0.59

POSITIVE LOGITS

nam

1.48

orously

1.12

ned

1.05

eless

1.01

borgh

1.00

ilitary

0.97

entary

0.94

ukong

0.92

elist

0.92

emn

0.90

Activations Density 0.003%

references to the name "nam" combined with specific surrounding context

occurrences of the word "nam" and possibly variations related to names or important figures

No Comments

No Known Activations