INDEX

Explanations

words related to suspicion or being suspected of something

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "suspected."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

idge

-0.74

rx

-0.72

skill

-0.68

reen

-0.67

bard

-0.66

ajo

-0.65

psey

-0.65

ammy

-0.65

arp

-0.65

ighth

-0.64

POSITIVE LOGITS

 suspect

0.78

 culprit

0.75

 suspects

0.71

 arson

0.68

iary

0.67

 mishand

0.66

 innocent

0.65

 Innocent

0.65

 offenders

0.64

 guilty

0.63

Activations Density 0.015%

words related to suspicion or being suspected of something

occurrences of the word "suspected."

No Comments

No Known Activations