INDEX

Explanations

negative sentiments or consequences associated with various topics

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

 finish

-0.78

 standoff

-0.68

SOS

-0.63

 Dickinson

-0.62

âĸ¬âĸ¬

-0.62

 cous

-0.61

sym

-0.61

 Sierra

-0.60

 table

-0.60

 saga

-0.59

POSITIVE LOGITS

thirds

1.52

ply

1.14

eyed

1.12

year

1.12

sided

1.11

legged

1.09

minded

1.06

third

1.06

fif

1.06

arms

1.05

Activations Density 0.044%

negative sentiments or consequences associated with various topics

No Comments

No Known Activations