INDEX

Explanations

instances of the word "briefly" or similar adverbs indicating short durations of time

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

lins

-0.74

rite

-0.71

endi

-0.70

imately

-0.65

yers

-0.65

Reviewer

-0.64

ERAL

-0.62

consumer

-0.62

artisan

-0.61

atu

-0.61

POSITIVE LOGITS

 halted

0.98

 paused

0.90

 interrupted

0.89

 housed

0.88

 blinded

0.86

 disrupted

0.84

 detained

0.83

 suspended

0.81

 derailed

0.80

 blocked

0.79

Activations Density 0.034%

instances of the word "briefly" or similar adverbs indicating short durations of time

No Comments

No Known Activations