INDEX

Explanations

mentions of a specific keyword "Hel" with varying emphasis indicated by different activation values

oai_token-act-pair · gpt-3.5-turbo

references to health-related topics and organizations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 Saber

-0.70

Eag

-0.67

 selves

-0.63

 negatives

-0.63

 surrogate

-0.60

 Memories

-0.58

 Lowell

-0.58

EED

-0.57

rers

-0.57

æī

-0.55

POSITIVE LOGITS

pless

1.24

mut

1.20

ms

1.17

ped

1.09

ping

1.07

mand

1.04

ios

1.02

iop

0.99

met

0.98

ps

0.96

Activations Density 0.038%

mentions of a specific keyword "Hel" with varying emphasis indicated by different activation values

references to health-related topics and organizations

No Comments

No Known Activations