INDEX

Explanations

numeric thresholds in various contexts

oai_token-act-pair · gpt-3.5-turbo

references to various types of thresholds, particularly in relation to risk and limitations

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

ocate

-0.75

ulf

-0.73

orld

-0.70

oir

-0.68

ocation

-0.67

 Merc

-0.66

utra

-0.65

ocated

-0.64

ocal

-0.63

ONG

-0.63

POSITIVE LOGITS

 thresholds

1.16

 threshold

1.08

witz

0.83

 cutoff

0.79

limits

0.75

 clearance

0.74

ansas

0.74

 heights

0.74

posts

0.73

 limits

0.73

Activations Density 0.010%

numeric thresholds in various contexts

references to various types of thresholds, particularly in relation to risk and limitations

No Comments

No Known Activations