INDEX

Explanations

names of political figures and titles

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

 transformer

-0.65

 replacements

-0.64

 detection

-0.63

inement

-0.63

 plague

-0.62

spr

-0.61

 valued

-0.61

 unification

-0.61

 numbering

-0.61

 trig

-0.60

POSITIVE LOGITS

 Yeah

1.34

Yeah

1.32

 Alright

1.28

 Exactly

1.27

Hmm

1.19

Okay

1.15

 Okay

1.13

Alright

1.09

Wow

1.07

Uh

1.07

Activations Density 0.672%

names of political figures and titles

No Comments

No Known Activations