INDEX

Explanations

information related to time spans, durations, and sequences

oai_token-act-pair · gpt-3.5-turbo

phrases related to events occurring over multiple years, particularly involving absence and return

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

 HERE

-0.67

humans

-0.61

 âĢº

-0.60

esi

-0.58

Origin

-0.55

 Unique

-0.54

isible

-0.54

 nutshell

-0.54

 ALWAYS

-0.54

 nowadays

-0.54

POSITIVE LOGITS

 thereafter

0.94

 afterward

0.77

 unsuccessfully

0.75

sequently

0.69

Reloaded

0.63

 afterwards

0.62

 subsequent

0.61

 himself

0.61

sequent

0.61

ļé

0.59

Activations Density 2.174%

information related to time spans, durations, and sequences

phrases related to events occurring over multiple years, particularly involving absence and return

No Comments

No Known Activations