INDEX

Explanations

words related to brands or products

oai_token-act-pair · gpt-3.5-turbo

references to "auxiliary" or supporting elements

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

enance

-0.78

lessly

-0.76

lessness

-0.75

ANK

-0.75

acco

-0.73

NN

-0.71

WAR

-0.70

mins

-0.67

IFT

-0.66

WAY

-0.65

POSITIVE LOGITS

iliary

1.34

uries

0.90

hall

0.82

illary

0.81

aux

0.80

odus

0.79

aux

0.78

ite

0.77

quire

0.77

chant

0.75

Activations Density 0.007%

words related to brands or products

references to "auxiliary" or supporting elements

No Comments

No Known Activations