INDEX

Explanations

specific nouns related to outcomes or results

oai_token-act-pair · gpt-3.5-turbo

various types of results or outcomes in contexts related to processes or experiences

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

inventoryQuantity

-0.75

terday

-0.71

dding

-0.71

 Ples

-0.66

 anten

-0.65

esan

-0.63

dates

-0.62

atal

-0.62

hun

-0.62

include

-0.61

POSITIVE LOGITS

 usable

0.89

 indistinguishable

0.88

 predictable

0.88

 results

0.83

 pleasing

0.83

 predictably

0.82

 chaotic

0.80

 result

0.79

 remarkably

0.77

ãĥĩãĤ£

0.77

Activations Density 0.247%

specific nouns related to outcomes or results

various types of results or outcomes in contexts related to processes or experiences

No Comments

No Known Activations