INDEX

Explanations

phrases related to a series of events

oai_token-act-pair · gpt-3.5-turbo

references to sequences of events or actions

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

istan

-0.73

laws

-0.72

ossier

-0.71

zie

-0.70

nai

-0.70

estine

-0.68

ablishment

-0.68

ittens

-0.68

istani

-0.67

idge

-0.67

POSITIVE LOGITS

encies

0.77

series

0.74

 finale

0.74

 series

0.69

ãĥ¼ãĥ³

0.68

 è£ıè¦ļéĨĴ

0.68

ãĥı

0.68

Series

0.66

meal

0.64

 resistor

0.64

Activations Density 0.020%

phrases related to a series of events

references to sequences of events or actions

No Comments

No Known Activations