INDEX

Explanations

instances of the letter 'K' in various contexts

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 unlaw

-0.62

 Adin

-0.61

 dere

-0.60

Qur

-0.60

 Hubble

-0.59

 diapers

-0.59

wcsstore

-0.59

 celebr

-0.59

 behavi

-0.58

Ø©

-0.58

POSITIVE LOGITS

orea

1.18

ernel

1.11

eeper

1.03

EEP

1.02

laus

0.96

istani

0.94

rieg

0.94

enzie

0.93

regate

0.92

TOR

0.91

Activations Density 0.021%

instances of the letter 'K' in various contexts

No Comments

No Known Activations