INDEX

Explanations

phrases related to potential risks or negative consequences

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate potential risks or threats

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

 agent

-0.58

Sec

-0.55

Fem

-0.55

bab

-0.54

 Desk

-0.53

Pod

-0.52

 Agent

-0.52

 preced

-0.52

 grips

-0.52

 mascul

-0.51

POSITIVE LOGITS

VIDIA

0.74

etheless

0.73

IUM

0.66

Electric

0.63

urai

0.62

served

0.61

 ILCS

0.61

maxwell

0.61

CLR

0.61

electric

0.60

Activations Density 0.000%

phrases related to potential risks or negative consequences

phrases that indicate potential risks or threats

No Comments

No Known Activations