INDEX

Explanations

adverbs that indicate performance or success

oai_token-act-pair · gpt-3.5-turbo

phrases that express effectiveness or quality in context

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

sing

-0.77

sin

-0.72

psy

-0.67

isms

-0.66

rian

-0.65

shaw

-0.64

clerosis

-0.63

rop

-0.63

weak

-0.62

cipl

-0.62

POSITIVE LOGITS

IER

0.78

 suited

0.78

ior

0.77

awaru

0.68

nesota

0.67

 synergy

0.67

[+]

0.65

 outwe

0.65

 depends

0.65

itsu

0.63

Activations Density 0.184%

adverbs that indicate performance or success

phrases that express effectiveness or quality in context

No Comments

No Known Activations