INDEX

Explanations

terms related to "Na" or "Naomi"

oai_token-act-pair · gpt-3.5-turbo

mentions of sodium or related chemical terms

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

eering

-0.87

tips

-0.84

 Journals

-0.79

ttes

-0.75

dress

-0.73

papers

-0.69

lords

-0.69

bucks

-0.69

 Colossus

-0.68

rers

-0.68

POSITIVE LOGITS

uthor

1.02

iber

1.00

elson

0.97

Ã¯ve

0.96

omi

0.92

Äį

0.91

emonic

0.91

umann

0.88

ples

0.88

emon

0.87

Activations Density 0.015%

terms related to "Na" or "Naomi"

mentions of sodium or related chemical terms

No Comments

No Known Activations