INDEX

Explanations

instances of the word "to" used in a specific context or followed by certain types of verbs

oai_token-act-pair · gpt-3.5-turbo

occurrences of the phrase "to" followed by verbs in various forms

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

keeping

-0.69

 Palest

-0.69

TED

-0.64

boys

-0.64

yet

-0.63

holding

-0.63

leading

-0.62

horn

-0.62

prem

-0.62

Reviewed

-0.61

POSITIVE LOGITS

 realize

1.24

 realise

1.20

 resemble

1.20

 accumulate

1.19

 crumble

1.13

 lose

1.13

 dismantle

1.12

 develop

1.11

 formulate

1.11

 emerge

1.08

Activations Density 0.052%

instances of the word "to" used in a specific context or followed by certain types of verbs

occurrences of the phrase "to" followed by verbs in various forms

No Comments

No Known Activations