INDEX

Explanations

words related to police operations like "raid"

oai_token-act-pair · gpt-3.5-turbo

references to police raids and law enforcement actions

oai_token-act-pair · gpt-4o-mini

Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Negative Logits

 laure

-0.75

 Warp

-0.71

BuyableInstoreAndOnline

-0.69

 Elev

-0.63

 Reilly

-0.62

atu

-0.61

åħī

-0.61

ãĤ´ãĥ³

-0.59

ocon

-0.59

 Bachelor

-0.58

POSITIVE LOGITS

 raid

1.29

 raids

1.23

 raided

1.09

 operation

1.05

 netted

1.03

 raiding

1.00

Operation

0.93

raid

0.91

 seizure

0.88

 warrant

0.88

Activations Density 0.073%

words related to police operations like "raid"

references to police raids and law enforcement actions

No Known Activations