INDEX

Explanations

instances of the word "proof."

oai_token-act-pair · gpt-3.5-turbo 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

 livest

-0.85

cffffcc

-0.72

 contrace

-0.69

ideshow

-0.67

lished

-0.66

 Peninsula

-0.65

artney

-0.64

ernel

-0.63

ecycle

-0.63

Osc

-0.61

POSITIVE LOGITS

reading

1.34

read

1.12

reader

1.11

edly

0.98

ing

0.88

ingham

0.88

reads

0.85

uers

0.84

ysis

0.83

 proofs

0.82

Activations Density 0.027%

instances of the word "proof."

No Comments

No Known Activations