INDEX

Explanations

phrases related to providing instructions or information

oai_token-act-pair · gpt-3.5-turbo

prompts for actionable advice or activities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

enhagen

-0.80

thora

-0.63

neys

-0.60

 ISBN

-0.59

obyl

-0.59

nu

-0.58

Ã¨re

-0.57

 Drac

-0.57

mate

-0.55

okia

-0.55

POSITIVE LOGITS

æĸ¹

0.72

IVE

0.69

APD

0.68

ä»£

0.63

phrine

0.62

VER

0.61

åĳ

0.60

Submit

0.60

 willpower

0.59

DEM

0.59

Activations Density 0.028%

phrases related to providing instructions or information

prompts for actionable advice or activities

No Comments

No Known Activations