INDEX

Explanations

terms related to enabling and disabling features or options

oai_token-act-pair · gpt-3.5-turbo

terms related to enabling or disabling features and options

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.3.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

apest

-1.03

fare

-0.81

ivan

-0.81

orians

-0.80

von

-0.79

ortment

-0.79

bah

-0.78

rium

-0.75

alg

-0.75

dor

-0.75

POSITIVE LOGITS

 disabling

0.87

 automatic

0.87

Disable

0.86

 disable

0.86

 defaults

0.81

 tuning

0.80

 CONFIG

0.77

 confir

0.76

 installs

0.76

 autop

0.75

Activations Density 0.039%

terms related to enabling and disabling features or options

terms related to enabling or disabling features and options

No Comments

No Known Activations