INDEX

Explanations

instances of actions or events that have been ongoing or repeated over time

oai_token-act-pair · gpt-3.5-turbo

phrases indicating ongoing actions or processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 Monstrous

-0.66

 desk

-0.60

 congr

-0.59

 calculator

-0.59

 Sins

-0.59

 courtesy

-0.58

beer

-0.58

 Rite

-0.58

 Copyright

-0.56

 intel

-0.56

POSITIVE LOGITS

 steadily

0.88

 continuously

0.80

 progressively

0.79

 countless

0.75

anco

0.74

 consistently

0.72

 relentlessly

0.70

 endeav

0.70

 successive

0.68

 continually

0.65

Activations Density 0.367%

instances of actions or events that have been ongoing or repeated over time

phrases indicating ongoing actions or processes

No Comments

No Known Activations