INDEX

Explanations

references to research and its related terminology

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

twitch

-0.72

engers

-0.64

 Leilan

-0.63

 Remastered

-0.61

awar

-0.61

alone

-0.60

 retaliate

-0.59

 Nationals

-0.59

generic

-0.59

 signalling

-0.59

POSITIVE LOGITS

sonian

0.82

 psychology

0.77

 researching

0.75

ãĥ´ãĤ¡

0.75

 scholar

0.74

Ô

0.72

 RAND

0.71

ondo

0.71

tainment

0.70

taining

0.70

Activations Density 0.035%

references to research and its related terminology

No Comments

No Known Activations