INDEX

Explanations

complimentary or positive words and phrases

oai_token-act-pair · gpt-3.5-turbo

positive descriptors or sentiments, particularly variations of the word "nice."

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

ogens

-0.74

onics

-0.74

obia

-0.70

uilding

-0.69

eligible

-0.68

doms

-0.68

authorized

-0.67

rupal

-0.67

ingo

-0.67

rules

-0.64

POSITIVE LOGITS

 little

0.93

 touches

0.90

 touch

0.89

 sounding

0.86

 neat

0.86

 nice

0.82

 breeze

0.82

 fluffy

0.81

 smelling

0.80

 gesture

0.79

Activations Density 0.041%

complimentary or positive words and phrases

positive descriptors or sentiments, particularly variations of the word "nice."

No Comments

No Known Activations