INDEX

Explanations

repeated phrases or patterns indicating an ongoing action or situation

oai_token-act-pair · gpt-4o-mini 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

rites

-0.72

anim

-0.71

Anim

-0.70

WARE

-0.69

fml

-0.69

gnu

-0.68

reference

-0.66

lua

-0.64

Reference

-0.63

Prev

-0.63

POSITIVE LOGITS

erous

1.02

 board

0.84

 roles

0.82

 behalf

0.82

 responsibility

0.74

 responsibilities

0.73

 sqor

0.69

 obesity

0.66

 extra

0.66

 undue

0.66

Activations Density 0.024%

repeated phrases or patterns indicating an ongoing action or situation

No Comments

No Known Activations