INDEX

Explanations

terms related to feasibility, viability, and usability

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

bery

-0.74

zona

-0.71

upt

-0.71

enfranch

-0.70

nikov

-0.70

Interstitial

-0.69

angelo

-0.68

down

-0.68

roman

-0.67

hani

-0.66

POSITIVE LOGITS

 feasibility

1.10

 feas

0.75

ibl

0.71

ancial

0.70

 sketches

0.69

izons

0.68

OLOG

0.68

otyp

0.68

 studies

0.67

 viability

0.67

Activations Density 0.005%

terms related to feasibility, viability, and usability

No Comments

No Known Activations