INDEX

Explanations

phrases related to detecting or finding something specific or suspicious

oai_token-act-pair · gpt-3.5-turbo

terms related to detection and identifying threats or issues

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

mort

-0.92

nesia

-0.72

emate

-0.71

erer

-0.71

orld

-0.70

wrap

-0.69

minus

-0.68

sis

-0.67

udeb

-0.66

ĸļ

-0.66

POSITIVE LOGITS

 flaws

0.78

 tampering

0.76

 detect

0.76

 leakage

0.74

ively

0.74

 detection

0.73

 faults

0.72

 emitting

0.71

 anomalies

0.70

 detects

0.69

Activations Density 0.056%

phrases related to detecting or finding something specific or suspicious

terms related to detection and identifying threats or issues

No Comments

No Known Activations