INDEX

Explanations

specific terms related to auditing and auditoriums

oai_token-act-pair · gpt-3.5-turbo

references to official titles and roles related to auditing and oversight

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

wikipedia

-0.66

âĵĺ

-0.66

 edible

-0.62

history

-0.61

riv

-0.60

 understanding

-0.60

PK

-0.60

 Immigration

-0.59

DUI

-0.59

Ã±a

-0.58

POSITIVE LOGITS

ium

1.23

 auditor

1.15

 Auditor

1.02

rors

0.90

rics

0.87

atures

0.87

iosyn

0.86

atis

0.82

Reviewer

0.80

atically

0.80

Activations Density 0.016%

specific terms related to auditing and auditoriums

references to official titles and roles related to auditing and oversight

No Comments

No Known Activations