INDEX

Explanations

code snippets related to implementing functions that take in other functions as arguments

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ahon

-0.75

ATHER

-0.65

ieri

-0.64

hig

-0.63

ciation

-0.62

 Harlem

-0.60

okÃ©

-0.59

pper

-0.58

 Lans

-0.58

areth

-0.58

POSITIVE LOGITS

locked

0.77

locks

0.76

main

0.74

oslav

0.71

rans

0.69

raints

0.68

tered

0.66

forge

0.66

escent

0.66

ulent

0.66

Activations Density 0.017%

code snippets related to implementing functions that take in other functions as arguments

No Comments

No Known Activations