INDEX

Explanations

articles or mentions about troubled individuals or situations

oai_token-act-pair · gpt-3.5-turbo

references to individuals or entities experiencing difficulties or challenges

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

©¶æ

-0.84

ITNESS

-0.80

galitarian

-0.76

ensional

-0.76

ILA

-0.76

entric

-0.75

ingham

-0.75

imates

-0.74

othy

-0.74

utra

-0.73

POSITIVE LOGITS

 troubled

0.87

 plag

0.73

 troubles

0.72

cffffcc

0.64

 waters

0.62

nesses

0.62

 nerves

0.62

 enterprises

0.61

 Ares

0.60

 adolesc

0.59

Activations Density 0.014%

articles or mentions about troubled individuals or situations

references to individuals or entities experiencing difficulties or challenges

No Comments

No Known Activations