INDEX

Explanations

terms related to ongoing actions or processes

oai_token-act-pair · gpt-3.5-turbo

the term "continuing" or its variations, indicating ongoing actions or situations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

eah

-0.81

idden

-0.77

ogy

-0.73

ard

-0.72

Ã¨re

-0.72

arta

-0.71

liest

-0.71

udging

-0.70

lier

-0.70

oster

-0.69

POSITIVE LOGITS

 unab

0.78

 uninterrupted

0.75

 revers

0.73

 appropriations

0.73

 unexplained

0.72

 antioxid

0.72

 hostilities

0.71

 strugg

0.70

 developments

0.70

 unresolved

0.70

Activations Density 0.016%

terms related to ongoing actions or processes

the term "continuing" or its variations, indicating ongoing actions or situations

No Comments

No Known Activations