INDEX

Explanations

phrases related to providing clarification or additional information

oai_token-act-pair · gpt-3.5-turbo

instances of the word "To" followed by explanations or clarifications

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 nets

-0.68

 Appears

-0.61

dro

-0.59

 forg

-0.58

 diapers

-0.57

lot

-0.55

calling

-0.54

fires

-0.54

pic

-0.53

urgy

-0.53

POSITIVE LOGITS

ilet

1.45

 summarize

1.24

 illustrate

1.12

 reiterate

1.11

ppings

1.10

 complicate

1.10

pping

1.08

 clarify

1.06

 compensate

1.06

 commemorate

1.02

Activations Density 0.044%

phrases related to providing clarification or additional information

instances of the word "To" followed by explanations or clarifications

No Comments

No Known Activations