INDEX

Explanations

statements of opinion or assertions

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

figure

-0.70

pes

-0.69

velength

-0.65

 mathemat

-0.64

ourses

-0.61

kefeller

-0.61

 focal

-0.61

 obser

-0.60

awar

-0.60

imeters

-0.58

POSITIVE LOGITS

 goodbye

1.06

 hello

0.81

 unequivocally

0.68

 definitively

0.68

 nobody

0.68

rists

0.66

 there

0.66

 that

0.65

 confidently

0.64

âĶĢâĶĢâĶĢâĶĢ

0.64

Activations Density 0.022%

statements of opinion or assertions

No Comments

No Known Activations