INDEX

Explanations

instances of the word "high" and related terms indicating elevated levels or severity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.8.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

sov

-0.71

RAW

-0.69

emis

-0.69

ools

-0.69

Ò

-0.68

uel

-0.65

ruction

-0.62

orio

-0.61

vice

-0.60

irit

-0.60

POSITIVE LOGITS

 enough

1.11

 compared

1.00

(>

0.96

 proportion

0.92

 percentage

0.86

(~

0.83

 relative

0.82

(<

0.80

 percentages

0.75

 paced

0.75

Activations Density 0.072%

instances of the word "high" and related terms indicating elevated levels or severity

No Comments

No Known Activations