INDEX

Explanations

phrases describing a process or workflow

oai_token-act-pair · gpt-3.5-turbo

statements explaining processes or concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

uable

-0.73

 prized

-0.62

worthiness

-0.62

 contamin

-0.62

 risked

-0.61

inventoryQuantity

-0.60

 Newsp

-0.60

 overcame

-0.60

reon

-0.60

worth

-0.59

POSITIVE LOGITS

 follows

1.13

 straightforward

1.10

 simple

1.02

 summarized

1.00

 similar

0.95

:[

0.92

 identical

0.88

 simplest

0.87

 chronological

0.87

 analogous

0.85

Activations Density 0.319%

phrases describing a process or workflow

statements explaining processes or concepts

No Comments

No Known Activations