INDEX

Explanations

words or phrases related to the potential end or result of a situation

oai_token-act-pair · gpt-3.5-turbo

references to the concept of "outcome."

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

ondo

-0.74

yi

-0.74

 brakes

-0.71

ker

-0.71

aps

-0.69

andise

-0.67

king

-0.67

sis

-0.67

cer

-0.67

mans

-0.66

POSITIVE LOGITS

 outcome

1.06

 outcomes

0.94

bringer

0.84

 thereof

0.80

fulness

0.74

romeda

0.68

 probabilities

0.67

lace

0.66

 Yanuk

0.65

 Orche

0.65

Activations Density 0.013%

words or phrases related to the potential end or result of a situation

references to the concept of "outcome."

No Comments

No Known Activations