INDEX

Explanations

elements related to specific individuals or names, particularly those associated with organizations or movements

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

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

 thirds

-0.78

mble

-0.75

querque

-0.73

rums

-0.70

 Rebels

-0.69

teenth

-0.69

 Erica

-0.69

taboola

-0.67

leneck

-0.67

checks

-0.67

POSITIVE LOGITS

oint

1.11

OY

1.04

avascript

0.93

osit

0.92

EG

0.91

ealous

0.89

OE

0.88

JP

0.88

IT

0.87

NI

0.87

Activations Density 0.005%

elements related to specific individuals or names, particularly those associated with organizations or movements

No Comments

No Known Activations