INDEX

Explanations

names identified as Billy at high activation levels

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_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ional

-0.93

aries

-0.93

igating

-0.89

igated

-0.82

ATING

-0.82

ational

-0.82

eled

-0.81

igation

-0.79

arin

-0.79

yrinth

-0.79

POSITIVE LOGITS

cock

1.02

 Graham

0.85

cakes

0.81

Gib

0.79

 Idol

0.78

Dee

0.78

 Joel

0.77

bird

0.76

xtap

0.75

 Mitchell

0.75

Activations Density 0.017%

names identified as Billy at high activation levels

No Comments

No Known Activations