INDEX

Explanations

phrases related to investigations and analyses

oai_token-act-pair · gpt-3.5-turbo

terms related to investigative journalism and research reports

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

à¨

-0.69

ardless

-0.63

displayText

-0.60

 appar

-0.58

ãĥ¼

-0.57

oke

-0.57

chan

-0.56

 Himself

-0.55

Py

-0.55

 biologically

-0.55

POSITIVE LOGITS

 titled

0.80

 entitled

0.76

antage

0.72

wart

0.70

mbudsman

0.69

arthed

0.68

aghan

0.67

warts

0.65

hot

0.63

icter

0.63

Activations Density 0.245%

phrases related to investigations and analyses

terms related to investigative journalism and research reports

No Comments

No Known Activations