INDEX

Explanations

comparisons of different scenarios or options in terms of desirability or efficacy

oai_token-act-pair · gpt-3.5-turbo

phrases suggesting improvement or superiority in situations or actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

acy

-0.65

phabet

-0.62

latest

-0.61

 Contains

-0.61

iasm

-0.60

RANT

-0.58

adem

-0.57

shire

-0.56

alle

-0.56

racuse

-0.54

POSITIVE LOGITS

 than

1.31

 Than

0.97

than

0.96

 suited

0.80

served

0.76

 behaved

0.76

 handled

0.74

 arrang

0.73

 avoided

0.71

 solved

0.70

Activations Density 0.089%

comparisons of different scenarios or options in terms of desirability or efficacy

phrases suggesting improvement or superiority in situations or actions

No Comments

No Known Activations