INDEX

Explanations

terms related to a specific topic or organization, specifically the "BC" entity

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

mble

-0.88

ktop

-0.77

ners

-0.67

paces

-0.66

"\

-0.65

glers

-0.64

stice

-0.64

geant

-0.64

zzle

-0.64

izational

-0.63

POSITIVE LOGITS

BG

0.93

rypt

0.90

ORN

0.89

IJ

0.89

OTT

0.86

oreal

0.85

orne

0.84

ASE

0.84

 Hydro

0.83

oin

0.81

Activations Density 0.017%

terms related to a specific topic or organization, specifically the "BC" entity

No Comments

No Known Activations