INDEX

Explanations

terms related to legal cases, historical events, and public figures

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

iously

-0.72

ãĤ®

-0.67

binary

-0.67

uni

-0.67

etsk

-0.66

fan

-0.65

dot

-0.65

iland

-0.64

illian

-0.64

secure

-0.63

POSITIVE LOGITS

 Representatives

0.85

 Directors

0.82

 Fame

0.81

 Commissioners

0.72

 Excellence

0.70

 Lords

0.70

 Ministers

0.68

 Roses

0.68

 Inquiry

0.67

 Governors

0.67

Activations Density 0.601%

terms related to legal cases, historical events, and public figures

No Comments

No Known Activations