INDEX

Explanations

phrases related to personal experiences and characteristics

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

 VIDEOS

-0.97

onies

-0.79

iasm

-0.72

cation

-0.71

mares

-0.71

iths

-0.69

seys

-0.69

uden

-0.69

 interfaces

-0.66

encies

-0.66

POSITIVE LOGITS

 negotiator

1.05

 recipient

1.05

 beneficiary

1.05

 millionaire

1.04

 trustee

1.02

 believer

1.01

 businessman

1.00

 member

0.99

 proponent

0.99

 leader

0.98

Activations Density 1.364%

phrases related to personal experiences and characteristics

No Comments

No Known Activations