INDEX

Explanations

phrases related to conflicts or intense situations involving people and authorities

oai_token-act-pair · gpt-3.5-turbo

phrases related to criminal activities and legal issues

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

Not in Any Lists

No Comments

Negative Logits

FIR

-0.59

Hes

-0.58

 Logged

-0.57

 Bliss

-0.57

 IPCC

-0.56

Ri

-0.55

 dispatch

-0.55

Lol

-0.55

Hut

-0.54

 grav

-0.54

POSITIVE LOGITS

scoring

1.04

bys

0.97

through

0.93

flat

0.91

rule

0.89

along

0.86

down

0.86

uncle

0.85

oriented

0.83

luster

0.83

Activations Density 0.038%

phrases related to conflicts or intense situations involving people and authorities

phrases related to criminal activities and legal issues

No Comments

No Known Activations