INDEX

Explanations

punctuation marks, particularly periods

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 advis

-0.73

 chopping

-0.68

 levers

-0.68

 wise

-0.68

 mant

-0.61

 microphone

-0.60

 provoking

-0.59

 confidentiality

-0.59

 homework

-0.59

 purse

-0.58

POSITIVE LOGITS

Va

0.97

COM

0.86

com

0.85

AX

0.84

NET

0.83

0.82

MX

0.81

rex

0.80

MO

0.79

Activations Density 0.034%

punctuation marks, particularly periods

No Comments

No Known Activations