INDEX

Explanations

The main thing this neuron does is find names related to the word 'iman'

oai_token-act-pair · gpt-3.5-turbo

references to a specific individual, likely a notable public figure or business person

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

Temperature

-0.79

CAST

-0.76

REE

-0.75

Values

-0.66

swing

-0.64

dies

-0.63

ÃįÃį

-0.63

LLOW

-0.62

Score

-0.62

ORK

-0.62

POSITIVE LOGITS

ufact

1.06

ovic

1.01

ipal

0.96

iman

0.95

agement

0.95

jar

0.95

aneous

0.90

agements

0.90

uated

0.89

amiya

0.87

Activations Density 0.007%

The main thing this neuron does is find names related to the word 'iman'

references to a specific individual, likely a notable public figure or business person

No Comments

No Known Activations