INDEX

Explanations

symbols and formatting-related elements in the text

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

 Canaver

-0.76

 Reincarn

-0.72

 Predators

-0.69

 Alzheimer

-0.68

 Ethiop

-0.67

hei

-0.67

 unfocusedRange

-0.66

RSA

-0.66

 Flames

-0.64

 Meth

-0.62

POSITIVE LOGITS

ican

0.82

eling

0.71

pr

0.70

atin

0.69

ining

0.69

ixt

0.69

cru

0.67

etr

0.67

ftime

0.66

inning

0.66

Activations Density 0.006%

symbols and formatting-related elements in the text

No Comments

No Known Activations