INDEX

Explanations

letters or symbols within a specific pattern

oai_token-act-pair · gpt-3.5-turbo

sequences of numerical and alphanumeric identifiers, including license plates and codes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

hement

-0.93

ouble

-0.69

igi

-0.64

glers

-0.62

utical

-0.62

abwe

-0.60

ushi

-0.60

hig

-0.60

 interfered

-0.59

amac

-0.59

POSITIVE LOGITS

ÂłÂłÂłÂłÂłÂłÂłÂł

0.79

ÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂłÂł

0.77

ÂłÂłÂł

0.72

ÂłÂłÂłÂł

0.68

↵

0.63

ÂłÂł

0.63

Filename

0.62

Synopsis

0.60

pmwiki

0.56

Sensor

0.54

Activations Density 0.484%

letters or symbols within a specific pattern

sequences of numerical and alphanumeric identifiers, including license plates and codes

No Comments

No Known Activations