INDEX

Explanations

information related to a specific topic or subject, potentially with a focus on technical details or analysis

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

irms

-0.65

ãģĨ

-0.65

 ç¥ŀ

-0.61

 segment

-0.60

tained

-0.60

built

-0.59

pointer

-0.59

":["

-0.58

 Annotations

-0.57

ãģ®

-0.57

POSITIVE LOGITS

tons

1.33

 alas

1.13

withstanding

0.94

romeda

0.90

chers

0.88

 beware

0.84

owsky

0.84

hey

0.83

tery

0.83

cher

0.82

Activations Density 0.100%

information related to a specific topic or subject, potentially with a focus on technical details or analysis

No Comments

No Known Activations