INDEX

Explanations

names or terms related to particular individuals or groups

oai_token-act-pair · gpt-3.5-turbo

names and references related to "Jam" and associated figures

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

pse

-0.68

 Marketable

-0.66

 Nietzsche

-0.65

 Scientific

-0.64

chnology

-0.63

 drawn

-0.61

 condoms

-0.61

 needles

-0.60

 contraceptive

-0.60

 Clicker

-0.59

POSITIVE LOGITS

aji

1.07

iah

0.99

oon

0.98

monds

0.97

mond

0.93

pling

0.92

pered

0.92

ayne

0.91

ison

0.87

olla

0.85

Activations Density 0.015%

names or terms related to particular individuals or groups

names and references related to "Jam" and associated figures

No Comments

No Known Activations