INDEX

Explanations

percentage values

oai_token-act-pair · gpt-3.5-turbo

percentages in various 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

 shaped

-0.73

pecially

-0.71

bley

-0.69

zbollah

-0.68

 showc

-0.68

iod

-0.68

 satell

-0.66

 brace

-0.66

 compr

-0.63

 compan

-0.63

POSITIVE LOGITS

ABV

0.74

 Discuss

0.73

vantage

0.68

 Overall

0.67

iles

0.67

 Sheet

0.65

 Decay

0.65

payer

0.62

/-

0.62

 Wanted

0.61

Activations Density 0.033%

percentage values

percentages in various contexts

No Comments

No Known Activations