INDEX

Explanations

actions related to completing tasks

oai_token-act-pair · gpt-3.5-turbo

words related to the concept of completion or finishing tasks

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

elled

-0.80

gravity

-0.74

uli

-0.74

esis

-0.67

LI

-0.67

hare

-0.66

alogy

-0.65

ships

-0.64

arta

-0.64

resy

-0.62

POSITIVE LOGITS

 unfinished

0.89

 touches

0.79

teness

0.71

housing

0.66

 preparations

0.66

 Pradesh

0.66

ãĤ¤ãĥĪ

0.65

 fabrication

0.64

 finishing

0.64

CLASSIFIED

0.63

Activations Density 0.023%

actions related to completing tasks

words related to the concept of completion or finishing tasks

No Comments

No Known Activations