INDEX

Explanations

information related to news articles about incidents or crime scenes

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

++++++++++++++++

-0.68

riott

-0.67

 Birch

-0.66

âĸ¬

-0.65

Wr

-0.63

 Falcon

-0.62

ministic

-0.59

natureconservancy

-0.59

rence

-0.59

uate

-0.59

POSITIVE LOGITS

yssey

1.09

hiba

0.92

otomy

0.90

ocom

0.90

imus

0.85

olesc

0.83

oro

0.82

opy

0.82

omb

0.82

gyn

0.80

Activations Density 1.067%

information related to news articles about incidents or crime scenes

No Comments

No Known Activations