INDEX

Explanations

phrases related to collective actions or statements

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

luster

-0.72

ahime

-0.65

 arresting

-0.61

 strang

-0.61

 Fleming

-0.61

CLR

-0.59

istant

-0.58

cible

-0.58

plete

-0.58

 Gothic

-0.57

POSITIVE LOGITS

ocating

1.07

 alike

1.04

 agree

0.93

together

0.90

 kinds

0.86

 collectively

0.85

 smiles

0.79

 together

0.79

 agreed

0.77

 benefited

0.76

Activations Density 0.034%

phrases related to collective actions or statements

No Comments

No Known Activations