INDEX

Explanations

names of individuals

oai_token-act-pair · gpt-3.5-turbo

proper nouns, particularly names and titles related to individuals and their actions

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

jah

-0.90

pex

-0.85

rum

-0.84

ãĥ¼ãĥ³

-0.83

 Kush

-0.82

oha

-0.81

ptive

-0.78

nda

-0.77

zy

-0.77

xia

-0.76

POSITIVE LOGITS

 Livingston

0.93

 Hawking

0.77

 Stevenson

0.73

 lett

0.71

IRD

0.69

umbers

0.65

 Barcl

0.62

 Parish

0.62

kered

0.62

eding

0.61

Activations Density 0.025%

names of individuals

proper nouns, particularly names and titles related to individuals and their actions

No Comments

No Known Activations