INDEX

Explanations

phrases related to specific people or entities being involved in various activities

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

 millenn

-0.65

nova

-0.63

Pav

-0.59

flu

-0.59

 unchecked

-0.58

 testament

-0.57

thur

-0.56

perm

-0.56

launch

-0.56

press

-0.56

POSITIVE LOGITS

 therein

0.77

enza

0.75

imental

0.69

ACTION

0.69

ãĥĦ

0.67

rences

0.66

seys

0.65

in

0.64

ioned

0.64

IRT

0.64

Activations Density 0.925%

phrases related to specific people or entities being involved in various activities

No Comments

No Known Activations