INDEX

Explanations

statements related to opinions or uncertainties

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

 millenn

-0.76

rers

-0.61

 spotted

-0.61

 qualitative

-0.57

 nodd

-0.55

Foot

-0.55

meyer

-0.54

Topics

-0.54

 corner

-0.53

"]=>

-0.53

POSITIVE LOGITS

::::::::

0.99

-.

0.98

Pa

0.95

::::

0.90

_.

0.88

soon

0.83

,,,,

0.80

selves

0.79

."

0.77

hack

0.76

Activations Density 0.301%

statements related to opinions or uncertainties

No Comments

No Known Activations