INDEX

Explanations

quotes starting with the word "Every" that reflect various topics and contexts

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

olas

-0.75

phabet

-0.74

dict

-0.67

ees

-0.65

Jah

-0.60

Pru

-0.58

 Papers

-0.58

 Send

-0.58

 Graves

-0.58

 Prompt

-0.56

POSITIVE LOGITS

THING

1.58

 conceivable

1.37

 imaginable

1.28

where

1.25

 single

1.06

single

1.06

 inch

1.04

WHERE

1.04

 ounce

0.97

thin

0.96

Activations Density 0.302%

quotes starting with the word "Every" that reflect various topics and contexts

No Comments

No Known Activations