INDEX

Explanations

instances of deception, falsification, misrepresentation, and misleading statements

oai_token-act-pair · gpt-3.5-turbo

terms associated with deception and the act of falsification

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

aho

-0.72

optional

-0.72

soc

-0.69

 Infantry

-0.67

lov

-0.66

til

-0.66

Streamer

-0.66

tis

-0.66

areth

-0.66

bour

-0.66

POSITIVE LOGITS

 falsely

0.98

 deception

0.97

 misrepresent

0.97

 allegations

0.91

 extortion

0.90

 deceit

0.89

 omission

0.88

 accusations

0.87

 billing

0.85

 fals

0.85

Activations Density 0.266%

instances of deception, falsification, misrepresentation, and misleading statements

terms associated with deception and the act of falsification

No Comments

No Known Activations