INDEX

Explanations

names or terms related to a specific location or organization, potentially related to criminal activity

oai_token-act-pair · gpt-3.5-turbo

references to specific people, particularly those with the name "Pas" or its variations

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

ocating

-0.71

icably

-0.70

ysis

-0.66

ski

-0.66

ship

-0.65

 Drift

-0.64

arty

-0.63

Reviewer

-0.63

urst

-0.63

YING

-0.62

POSITIVE LOGITS

aurus

1.18

ync

1.10

pect

1.03

ques

0.99

daq

0.98

chal

0.97

que

0.96

coe

0.96

ocial

0.94

peed

0.94

Activations Density 0.056%

names or terms related to a specific location or organization, potentially related to criminal activity

references to specific people, particularly those with the name "Pas" or its variations

No Comments

No Known Activations