INDEX

Explanations

mentions of the word "Moss."

oai_token-act-pair · gpt-3.5-turbo

references to the term "Mossad" in various contexts

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

versions

-0.82

į

-0.78

IENCE

-0.77

Interstitial

-0.74

¹

-0.73

®

-0.70

cers

-0.67

nce

-0.67

realDonaldTrump

-0.66

iences

-0.66

POSITIVE LOGITS

berg

0.98

 Moss

0.95

creen

0.85

es

0.85

olini

0.81

ett

0.78

abee

0.74

 Lerner

0.72

osate

0.71

qv

0.71

Activations Density 0.012%

mentions of the word "Moss."

references to the term "Mossad" in various contexts

No Comments

No Known Activations