INDEX

Explanations

Then main thing this neuron does is find occurrences of the word "confidential" along with related context like "confidential information" and "confidentiality"

oai_token-act-pair · gpt-3.5-turbo

terms related to confidentiality and sensitive information

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

alis

-0.94

Ì¶

-0.90

annis

-0.85

ð

-0.79

ixels

-0.78

hao

-0.74

plex

-0.71

geist

-0.71

stru

-0.70

imus

-0.69

POSITIVE LOGITS

 confidential

1.15

 informant

1.07

 informants

1.02

 correspondence

0.89

idential

0.89

arial

0.85

 information

0.84

 documents

0.81

 confidentiality

0.81

 arrangements

0.78

Activations Density 0.021%

Then main thing this neuron does is find occurrences of the word "confidential" along with related context like "confidential information" and "confidentiality"

terms related to confidentiality and sensitive information

No Comments

No Known Activations